Lithium, as the lightest alkali metal, is widely used in many industries such as batteries, medicine, and ceramics due to its physical and chemical properties, including high specific heat capacity, low expansion coefficient, and the highest electrochemical activity redox potential^[1⇓⇓~4]. In recent years, due to the rapid development of the new energy industry, the global demand for lithium resources has increased sharply. According to data from the U.S. Geological Survey (USGS)^[5], the global consumption of lithium metal reached 134,000 tons in 2022, with a growth rate of 41%. Among this, the battery industry accounts for the largest proportion of global lithium resource consumption (see Figure 1). It is estimated that by 2040, the demand for lithium will exceed its production by 8 times^[6].

Lithium resources are widely distributed across various countries globally. According to the 2023 USGS data^[5], the global identified lithium resource reserves (calculated as Li₂O) amount to approximately 98 million tons, mainly distributed in Bolivia (21 million tons), Argentina (20 million tons), the United States (12 million tons), Chile (11 million tons), Australia (7.9 million tons), China (6.8 million tons), and other countries. The confirmed lithium resources of different countries worldwide are shown in Figure 2.

Lithium resources mainly exist in two forms: one is in lithium-bearing ores, such as pegmatite-type lithium deposits and sedimentary-type lithium deposits; the other is in ionic form in brine resources, such as salt lake brines, seawater, deep brines, and oilfield brines. Although the extraction technology for the former is more mature, over 60% of global lithium resources are found in salt lakes and seawater, particularly in brines located in Chile, Bolivia, Argentina, China, and the United States^[7]. Therefore, there is an urgent need to develop efficient liquid lithium resource extraction technologies.

The precipitation method is a technique that separates the target component from a solution through chemical reactions, and it is the simplest method for extracting lithium from brine, which has been widely used in large-scale industrial lithium extraction^[9], but it is only suitable for brine with a magnesium to lithium mass ratio below 8. The Atacama Salt Lake in Chile, South America, is a typical representative of low-magnesium-to-lithium-ratio salt lakes. The typical process flow for lithium extraction is: after most of the sodium chloride and potassium salts are removed by pre-solar evaporation, the brine is mixed with a calcium-containing solution to form gypsum, removing sulfate from the brine. Continued solar evaporation increases the mass fraction of Li⁺ to about 4.3%, then slaked lime is used to raise the pH of the brine to around 11, removing most of the magnesium ions and sulfate. Na₂CO₃ is subsequently used to remove residual calcium and magnesium from the brine. Finally, the mother liquor, after being treated with Na₂CO₃ and filtered under heating, is filtered and dried to produce lithium carbonate^[10]. In recent years, researchers have developed several new types of precipitants. Zhang et al.^[9] synthesized a novel sodium magnesium metasilicate precipitant for the separation of lithium and magnesium in high-magnesium-to-lithium-ratio brine, as shown in Figure 4. This precipitant can achieve a 86.73% Li⁺ recovery rate and a 99.94% Mg²⁺ removal rate.

Liu et al.^[11] synthesized an Al/Na₂SO₄ composite precipitant capable of forming lithium precipitates in the form of Li₂Al₄(OH)₁₂SO₄∙xH₂O, but the presence of magnesium is not conducive to the precipitation of lithium. When the magnesium-to-lithium ratio reached 20, the lithium precipitation rate decreased from 89.2% to 54.7%. Liu et al.^[12] proposed a method for activating Li₃PO₄ to induce carbonate precipitation, successfully preparing active lithium phosphate with exposed high-surface-energy (110) facets. This precipitant significantly reduced the temperature required for the reaction (from 90 ℃ to 30 ℃), thereby lowering the energy consumption of the process.

Solvent extraction method, with advantages such as low cost, simple operation, and efficient separation, is widely used for the separation of Li^+[13,14] from various salt lake brines. The key to this method lies in the selected extraction agent system, which can be specifically categorized into crown ether system extractants, β-diketone system extractants, phosphorus-containing organic extractants, and ionic liquid system extractants, etc.^[15].

The adsorption method, due to its simplicity, environmental friendliness, and cost-effectiveness, is widely used for the separation of different ions in liquids^[29,30]. During the extraction process, lithium is selectively adsorbed and then desorbed by the solution, thereby separating it from coexisting ions^[31]. Currently, the main adsorbents for extracting liquid lithium resources are aluminum salt adsorbents and lithium ion sieve adsorbents. Aluminum salt adsorbents have a simple process, are environmentally friendly, and do not require stringent magnesium-to-lithium mass ratios in brine; they are the most mature and the only adsorbents that have been industrially applied so far^[32]. However, their maximum theoretical adsorption capacity is not high (6.1 mg/g)^[33], and the adsorption effect of lithium in solutions with low lithium concentrations is poor, which limits their application to some extent.

Lithium ion sieves have attracted extensive research from scholars at home and abroad since their discovery due to their high chemical stability, strong lithium adsorption capacity, unique chemical structure, and high selectivity. Lithium ion sieve adsorbents, which possess Li⁺memory, are divided into two types: manganese-based lithium ion sieves and titanium-based lithium ion sieves, and can be used for extracting lithium from salt lakes or seawater. The preparation process of lithium ion sieves is shown in Figure 6: the target Li⁺is introduced into inorganic compounds to form a lithium ion sieve precursor; without changing the crystal structure of the precursor, Li⁺is extracted with an acid washing solution, resulting in a lithium ion sieve adsorbent with a pore structure that matches the size of Li^{+ [7,34]}. The crystal vacancies formed in the lithium ion sieve after acid washing can accommodate ions with a radius equal to or smaller than that of the target Li^{+ [35]}. Compared with other metal ions, Li⁺has the smallest hydrated ionic radius, which is the reason why lithium ion sieves can selectively adsorb Li⁺and is also key to the synthesis process of lithium ion sieves.

In lithium ion sieves of the manganese system, due to the changes in the valence states of manganese (Mn⁴⁺, Mn³⁺, Mn²⁺), several lithium manganese oxide phases with different crystal structures can be synthesized^[36]. Among them, the precursors with higher Li⁺ adsorption capacity mainly include three types: LiMn₂O₄, Li_1.33Mn_1.67O₄, and Li_1.6Mn_1.6O₄^[37]. Manganese-based ion sieves have the drawbacks of poor stability and high dissolution loss rate of manganese^[38]. Therefore, researchers reduce the dissolution loss of manganese and improve the stability of the ion sieve by doping other metals into the precursors. Cao et al.^[39] synthesized a chromium-doped manganese-based lithium ion sieve Li_1.6Mn_1.6-xCr_xO₄ through hydrothermal reaction. When x is 0.016, the material's adsorption capacity for Li⁺ is 31.67 mg/g, and the dissolution rate of Mn is 2.1%. After 20 cycles in salt lake brine, the dissolution rate of manganese in the material was only 0.35%. Wang et al.^[40] prepared a zirconium-coated lithium ion sieve by coating zirconia on the surface of Li_1.6Mn_1.6O₄. After coating, the dissolution loss rate of Mn²⁺ in the lithium ion sieve decreased from 0.89% to 0.349%.

Compared to manganese dioxide, titanium dioxide exhibits better stability in acidic media and, due to its environmentally friendly nature and ease of separation from aqueous solutions^[7], titanium-based lithium ion sieves have garnered increasing attention. Titanium-based lithium ion sieves are primarily divided into two types: H₄Ti₅O₁₂ (spinel structure) and H₂TiO₃ (layered structure)^[34,41]. Although the Ti⁴⁺ in H₄Ti₅O₁₂ has good chemical stability, the theoretical adsorption capacity of H₄Ti₅O₁₂ is 63.77 mg/g, which is much lower than that of H₂TiO₃ (142.9 mg/g)^[42]. Gu et al.^[43] used CH₃COOLi∙2H₂O instead of Li₂CO₃ as a lithium source, mixed with TiO₂ for calcination to synthesize Li₂TiO₃. During the early calcination stage, the melting of CH₃COOLi∙2H₂O (66 ℃) formed a liquid-solid phase, significantly improving the mixing of CH₃COOLi∙2H₂O and TiO₂. The substantial heat and gas released during dehydration accelerated the nucleation process, effectively inhibiting agglomeration, making the particle size of Li₂TiO₃ smaller and greatly reducing the ion exchange resistance. The adsorption capacity of the lithium ion sieve synthesized by this method reached 24.97 mg/g in brine from a salt lake in western Taiwan. Sun et al.^[44] prepared nanoscale lithium ion sieves H₂TiO₃-4 (HTO-4) using a templating method. This material exhibited excellent hydrophilicity and adsorption performance, with a maximum adsorption capacity of 56.03 mg/g. Yu et al.^[45] developed a titanium-based lithium ion sieve monolithic adsorbent using 3D printing technology, where the content of the titanium-based lithium ion sieve was as high as 75%. The mechanical strength of this adsorbent was 51.0 MPa, and the adsorption capacity in simulated brine could reach 12.3 mg/g within 4 hours.

Membrane separation is an emerging liquid lithium extraction technology, with advantages such as high separation efficiency, continuous operation, and strong production stability^[32]. The essence of membrane separation is the selective migration of ions, which can be specifically divided into nanofiltration (pressure-driven)^[46], electrodialysis (electric potential-driven)^[47], and a combination of nanofiltration and electrodialysis.

Pelletizing is one of the most commonly used methods for shaping lithium ion sieve powder adsorbents. Using this method, adsorbents with high mechanical stability and good water permeability can be obtained. Lin et al^[58]used acid and alkali resistant polyvinyl chloride (PVC) as a binder and polyethylene glycol (PEG-6000) as a pore-forming agent to prepare porous PVC-HTO granules. This material has an adsorption capacity of 12.84 mg/g at 328.15 K, with an equilibrium time of 12 h, and the decrease in adsorption capacity after 5 cycles is less than 2%. Yang et al^[59]synthesized a series of HTO-PVC-x(x = 10, 15, 20, 25, and 30) lithium ion sieves with different PVC contents using the anti-solvent method. Among them, HTO-PVC-15 exhibits the best adsorption performance, with an adsorption capacity of 22.34 mg/g at 25 ℃, reaching 96% of its adsorption capacity within 2 h. Zhao et al^[60]used poly(vinyl alcohol) (PVA) as a binder and Silane glycidoxy propyltrimethoxysilane (GPTES) as a coupling agent to form PVA/LMO-1 spherical composite materials from Li_1.33Mn_1.67O₄(LMO-1) powder. The specific surface area of these granular PVA/LMO-1 composites is 20.09 m²/g, and their Li⁺adsorption efficiency in Puguang salt lake is 93.5%.

Magnetic separation is an effective and simple method for separating particles from liquids or other fluids. Magnetic particles can be separated from the solution through a magnetic field, and the separated particles can be reused after removing the adsorbed substances. This method has advantages such as being non-toxic, low-cost, and easy to synthesize, and it is widely applied in fields like catalysis, water purification, and biomedical applications. Kim et al.^[61] reported a method of preparing magnetic lithium adsorbent composites by growing Fe₃O₄ on the precursor of a lithium ion sieve. The adsorption capacity of this method for lithium was 6.84 mg/g in LiCl buffer solution and 12.2 mg/g in poor quality concentrated seawater. After 6 cycles of adsorption-desorption tests, the material's adsorption capacity for lithium remained above 86%. Xue et al.^[62] synthesized Fe₃O₄-doped magnetic manganese-based lithium ion sieve HMO/FO using a hydrothermal co-precipitation method. The adsorption capacity of this adsorbent for lithium was 29.33 mg/g, and after 5 cycles of adsorption-desorption, the sample's Li⁺ adsorption capacity could remain at 19.07 mg/g. In recent years, Wang et al.^[63] synthesized iron-doped lithium titanate (Fe-Li₂TiO₃) through a solid-state doping reaction, which, after acid treatment, formed an iron-doped lithium ion sieve (Fe/Ti-x(H)). The magnetization adsorption efficiency and lithium adsorption capacity of the resulting solid powder were 96% and 53 mg/g, respectively.

Film formation is an effective way to improve the practicality of ion sieve powders in actual lithium recovery applications^[64]. Nanofibers, due to their high surface area and porosity, ultra-fine fiber structure, and excellent mechanical structural properties, are receiving increasing attention^[65⇓~67], and they have broad prospects in the field of ion sieves. Lawagon et al.^[68] used different polymers, including polyvinyl chloride (PVC), polysulfone (PSf), polyvinylidene fluoride (PVDF), and polyacrylonitrile (PAN), to synthesize a series of hybrid nanofibers containing H₂TiO₃ and polymers through electrospinning. The results showed that H₂TiO₃/PAN nanofibers had the best adsorption performance for Li⁺, with an adsorption capacity of 72.75 mg/g. Park et al.^[69] prepared activated lithium ion sieve/polysulfone nanofiber composite membranes through electrospinning, thermal annealing, and acid washing. Although the surrounding matrix blocked some of the adsorption sites of LIS, leading to a decrease in the adsorption capacity of the prepared lithium ion sieve nanofiber membrane compared to Li_0.67H_0.96Mn_1.58O₄ ion sieve particles, the dissolution loss of Mn was smaller in subsequent cycle tests.

Sun et al.^[70] prepared another polymer ionic sieve membrane by phase inversion of a mixed casting solution of H₄Mn₅O₁₂ ionic sieve and PVDF. The adsorption capacity of this H₄Mn₅O₁₂/PVDF membrane for Li⁺ can reach 27.8 mg/g. After 6 adsorption-desorption cycles, its adsorption performance can still be maintained above 90%. Inspired by the 2D MXene/1D CNF flexible film (MC), Wang et al.^[71] synthesized a hydrogen-bond-induced flexible hybrid film HMO@MC. This flexible film can achieve an HMO loading rate of 75% by mass, with a maximum lithium ion adsorption capacity of 21.39 mg/g. In addition, the MC film substrate also plays a role in intercepting interfering ions during the Li⁺ migration process through its surface functional groups, making the HMO@MC film have extremely high Li/Na and Li/Mg selectivity, as well as excellent lithium adsorption performance (18.23 mg/g HMO, 12 h) in simulated seawater.